This study investigates the effect of structural modification actions on the existing stormwater collecting system in Eastern Tehran to increase the hydraulic capacity and create suitable conditions for the passage of runoff in the critical points of the canal. First, the hydraulic conditions within the stormwater collecting system were simulated using the Stormwater Management Model (SWMM) model before/after the modification to investigate the rehabilitation results. Three critical locations along the main canal were recognized as the most vulnerable points. Then, based on field visits and brainstorming sessions, rehabilitation methods were presented, and three practical solutions, including canal deepening, canal widening, and their combination, were investigated for each. Then, local investigating based on the rehabilitation alternatives for each critical location was conducted using the HEC-RAS. Finally, the SWMM model was used again to evaluate the overall operational performance of the stormwater collecting system after the rehabilitation. The results revealed that it is necessary to implement two alternatives of deepening and widening to provide adequate transmission runoff capacity during rainfalls with various return periods. More specifically, the localized redesign of the eastern flood diversion canal had an acceptable improvement in reducing flooding problems so that for floods with a return period of 10 years, the number of node flooding dropped from 4 to 0, inundated areas from 17% to 0, and the overflow volume from (10–45) to 0. Moreover, the proposed local rehabilitation reduced the overflow volume from (30–65), (43–74), and (70–92) in the status quo to (4–12), (11–27), and (24–36) for rainfall with 25, 50, and 100-year return periods.

  • A centralized-decentralized operational performance assessment approach was proposed.

  • SWMM model was performed to identify critical points in the stormwater collection system.

  • Critical canal segments were detected based on hydraulic simulation.

  • Localized redesign alternatives were investigated using with HEC-RAS model.

  • A practical and straightforward approach was proposed for stormwater collection systems rehabilitation.

Graphical Abstract

Graphical Abstract
Graphical Abstract

River floods, especially in urban areas, are one of the most catastrophic phenomena, threatening developed and developing countries with a large share of financial and human losses (Hua et al. 2020). Urban flood is caused by the lack of capacity of the stormwater collecting systems against rainfall events in urban areas. Two main factors can amplify the intensity of the damage caused by this phenomenon. One of them is population growth and rapid urbanization, where human activities in the natural hydrological cycle, such as the conversion of natural basins to impermeable surfaces, removal of vegetation cover, soil sealing, illegal occupation of flood-prone areas, construction in natural routes, and artificial reservoirs developments, weaken the infiltration process and reduce evapotranspiration (Silva et al. 2014; Movahedinia et al. 2019). The second factor is extreme rainfall exacerbated by climate change, which plays a crucial role in intensifying and accelerating the hydrological cycle and might change the magnitude and frequency of rainfall and floods caused by changes in rainfall pattern (Kvočka et al. 2016; Dong et al. 2017; Patel et al. 2017). This factor impacts stormwater occurrence, volume, and peak flow in arid and semi-arid regions where rainfall usually occurs for a short time but with high intensity (Lázaro et al. 2016; Al-Zahrani et al. 2017).

Therefore, urban flood control as a fundamental and necessary measure to reduce financial, environmental, and human losses becomes the municipalities' priority (Xie et al. 2017; Nguyen-Tien et al. 2018). In order to achieve appropriate recovery plans and effective management plans to reduce the volume of runoff and stormwaters peak flow, as well as forecasting and warning floods systems, various structural modifications, non-structural measures, and combined alternatives have been implemented. These methods have significantly reduced the adverse effects of stormwater and floods in vulnerable urban regions (Lee & Kim 2018). The non-structural measures include but are not limited to urban master planning laws and regulations review, develop efficient flood warning and forecasting systems, optimal operating procedures during flood events, and regular inspections and maintenance activities of the stormwater collecting systems (Mel et al. 2020a). However, the proper implementation of the non-structural solutions is highly dependent on the operators, managers, and inspection personnel's skills, engineering judgment, knowledge, and experience, which increases the uncertainty of successful non-structural measures implementation (Wu et al. 2020).

However, structural approaches as a strategic, fundamental, and sustainable strategy to control urban flood is an inevitable principle (Gupta et al. 2018). Structural approaches include redesigning the drainage canals and constructing various flood protection facilities such as ditches, levees, diversion canals, spurs, culverts, tunnels, detention reservoirs, and rain ditch (Coombes & Downs 2015; Matos Silva & Costa 2016), position and capacity optimization of storage tanks (Tao et al. 2017), pumping stations implementation in connection with storage tanks for receiving and transferring urban runoff (Hsu et al. 2013). Patel et al. (2017) reported that due to protection measures such as dams and embankments as well as retaining walls implemented in Surat, India, the city's main areas are safe against floods with a flow rate of 14,430 m3/s. In another study, urban flood control and mitigation in Seoul, Korea, was investigated due to the redesigning of decentralized detention reservoirs and their relationship to the new input/output management. The results showed a significant reduction in flood volume of about 75 and 65% compared to the flood events in Seoul in 2010 and 2011, respectively (Lee & Kim 2017). Tobio et al. (2015) assessed the application of the Stormwater Management Model (SWMM) in determining the optimal physical design properties of an established low-impact development (LID) system treating road runoff. This study's results show that reducing the storage volume's original ratio to the facility's surface area and depth by 25% could match the satisfactory performance efficiency achieved in the original design. Lord et al. (2021) investigate the effect of the redesign of the stormwater collection canal system based on flood exceedance probability by integrating the optimization model of ant colony optimization and SWMM simulation models in a real test case in Tehran, Iran. Tansar et al. (2022) evaluates the performance of urban drainage system under different spatial placement strategies of LID to understand how urban flood dynamics of drainage system changes at catchment and local scales in a real test case in China. The study's findings confirmed the significant impact of the placement location of LID on stormwater collecting so that the placement of LID facilities nearer to the flooded locations maximizes the benefits in terms of flood reduction and also reduces the probability of transferring hydraulic load to other parts of the system. Yin et al. (2021) proposed an improved ant colony optimization method to enhance the stormwater collecting systems, where the optimization efficiency is enhanced by using an approximate design solution identified by the engineering design method. Their results show that the proposed method can identify design solutions with significantly improved efficiency and solution practicality compared to the traditional design approach, with advantages being more prominent for more significant stormwater collecting design problems.

Combining structural and non-structural methods is also recommended in different studies as effective measures for flood mitigation and recovery (Bertilsson et al. 2019). These methods include the optimal design of the structures of drainage canal networks (Farzin & Valikhan Anaraki 2020; Lord et al. 2021), flood transmission and control structures (Erdbrink et al. 2014; Seifert & Moore 2018), and real-time automated systems for flood drainage pumping stations (Wang et al. 2019). Other measures include the multi-objective optimization of reservoir management and operation (Meng et al. 2019; Yin et al. 2021), determining sustainable development policies, and planning structural actions (Nones 2015).

In light of the above, various structural, non-structural, and combined alternatives have been developed based on the topographic conditions of the area and technical, social, economic, and environmental considerations. However, social constraints, the dense texture of populated cities, and the lack of cooperation of society in implementing and maintaining the LIDs and modern storage facilities, especially in developing countries, are the main constraints that have made their use challenging. Therefore, the structural measures, including modification of the flood control canals and the hydraulic structures located in their paths, which are responsible for transporting the main load of the stormwater, have been introduced as the front line of rehabilitation alternatives from the local municipalities and cities (Rodríguez et al. 2000; Juan et al. 2020). The optimal redesign of the stormwater collecting canal networks using optimization techniques (Lord et al. 2021; Yin et al. 2021) is noteworthy in the rehabilitation projects. Nonetheless, the possibility of redesigning the entire canal network faces financial constraints in many dense urban areas (Mel et al. 2020b). Therefore, this study tried to present a practical-oriented approach based on (i) determining problematic canal segments and critical nodes upon the performance evaluation of the stormwater collecting system using the SWMM simulation model; (ii) local redesign of recognized canal segments through the canal widening, deepening, or combined form of both using the Hydrologic Engineering Center's River Analysis System (HEC-RAS) model; and (iii) performance evaluation of the rehabilitated canal systems in reducing the inundated areas, the overflow volume and the number of node flooding using SWMM simulation model. It is worth noting that the experiences of conducting the proposed method in the study area show that the local decision-makers and city managers were satisfied because their opinions and perspectives for finalizing the critical points – gathered during field visits, brainstorming session, and interviews – were included in the final renovation plan. It is worth noting that, to maintain the objectivity of the study and to deal with the subjectivity of the experts, 35 experts with more than 15 years of experience in design, construction, project management, technical analysis, environmental studies, and urban planning and development were included for the decision making process. These experts review the preliminary and final results and check any other possible alternatives during the study. In light of the above, the high number of experts with different areas of expertise help this study avoid bios with qualitative data from the experts' opinions. Besides, by including the different players in the methodology – i.e., the experts with different disciplines – the simulation models' role as a reliable decision support system was highlighted and accepted.

In light of the above, successful implementation of the objectives of the present study rely on the simulation results. In general, urban municipalities are generally interested in using simulation models to estimate the costs and benefits of adopting structural approaches before allocating funds and resources (Movahedinia et al. 2019). So far, many computer models such as HEC-RAS, HEC-HMS, MIKE11, MIKE SWMM, MIKE FLOOD, FLOW-3D, and InfoWORKS have been used to simulate river floods and the main surface water collecting systems. One of the most widely used models in hydraulic and hydrological studies of urban hydrology is the SWMM model. Also, the HEC-RAS simulation model has been used to evaluate the channelized and setbacks approaches in river flood control. Juan et al. (2020) used the Vflo model for hydrological simulation and the HEC-RAS model for hydraulic analysis of canals and flood extent in two urban areas under two scenarios with return periods (RPs) of 10 and 100 years and three intervals. They showed that structural approaches have been particularly effective for cities experiencing rapid population growth in better understanding long-term performance towards river flood risk. Silva et al. (2014) used the HEC-RAS mathematical hydrodynamic model with zoning floods in the urban area of Rio dos Cerdos in Brazil for events with an RP of 2, 5, 10, 20, 50, and 100 years. The results showed that roads and buildings around the bridge, as well as the confluence of river tributaries, are more vulnerable to flood risk. This study also demonstrated the potential of using HEC-RAS and geographic information system (GIS) models with high-resolution spatial data in zoning areas with high flood risks. To evaluate the performance of MIKE11 and HEC-RAS hydraulic models in river flood risk prediction studies in Malaysia, Alaghmand et al. (2012) concluded that the HEC-RAS model is more capable of flood risk zoning than the MIKE11 model based on four criteria, namely: data validity, available output results, usability, and availability. Moreover, the application of the HEC-RAS model has been investigated by Darajatun et al. (2020), Farooq et al. (2019), Ardıçlıoğlu & Kuriqi (2019), Carling et al. (2010), Pinar et al. (2010), and Ranaee et al. (2009).

Methodology

This study aims for flow capacity improvement and local rehabilitation schemes for flood control canals in these areas. The flowchart of this study includes (Figure 1):
  • surface runoff simulation in some of the canals in the study area and determining the inundation points (critical points) in the old and densely populated areas within the study area using the SWMM model under scenarios of rainfall with specific RPs;

  • developing a hydraulic model for the flow in the selected canals using HEC-RAS to simulate the flow in critical bridges accurately;

  • considering the rehabilitation action plans, including deepening and widening the canal reach, as well as their combination, based on brainstorming sessions;

  • investigating the hydraulic behavior of the localized redesign alternatives of the selected canals in HEC-RAS;

  • and performance evaluation of the entire stormwater collecting system after the localized redesign of the selected canals for precipitation with the specific RP in SWMM.

Figure 1

Research flowchart consisting of (a) performance simulation of stormwater collecting system using SWMM and (b) hydraulic simulation using HEC-RAS.

Figure 1

Research flowchart consisting of (a) performance simulation of stormwater collecting system using SWMM and (b) hydraulic simulation using HEC-RAS.

Close modal

Study area

The main runoff collection network in the east of Tehran was selected as the study area in the research, between the latitude of 35° 36′ 24″ to 35° 48′ 41″ north and longitude of 51° 24′ 11″ to 51° 38′ 30″ east. The main canal is about 207 km. Most of the main canals in this network are north-south-directed, with a moderate slope in the city's general slope, directing runoff from mountainous and urban areas to the main collector outside the city. Figure 2 shows the location of the area to the municipality of Tehran. To collect, transfer and control runoff in this network, only the traditional concrete canal system and other affiliated structures such as drop and transverse structures are located at the intersection with the area's main streets. Field studies have been observed as hydraulic constraints based on the study area maps. For example, due to the rape of residential areas and construction in the canal domain, the cross-section area has gradually declined compared to upstream sections, which indicates a decrease in hydraulic potential in various parts of the network. In some places, especially local narrowing within the cross-bridge range at the intersection with the main streets, the cross section area has reduced up to 50% over the upstream sections. In addition, the turbulence and erosion of canal beds in various sections due to the roughness of the bed and the non-standard conduct of the wall and floor rehabilitation processes along some parts of the network have been observed.
Figure 2

Study area, located in the eastern region of the Tehran metropolis.

Figure 2

Study area, located in the eastern region of the Tehran metropolis.

Close modal

The required information for developing the simulation models of SWMM and HEC-RAS included but was not limited to (i) study area's land use and cadaster maps to extract property's metes-and-bounds of the sub-catchments; (ii) topographic map and climatology and hydrological data; (iii) study area's characteristics such as soil type, slope, and other details; (iv) the drainage canal network maps in AutoCAD format; (v) the measured rainfall data extracted from the city's data warehouse; (vi) measured runoff information during the rainfall events by the on-site flow meter stations stored in the data loggers; and (vii) cadaster and land-use maps. It is worth noting that the detailed information mentioned above was gathered from different offices of the Tehran City Co. (Tehran Municipality).

SWMM simulation model

As a dynamic model, SWMM simulates the quality and quantity of runoff continuously in urban areas for a particular event using the Manning and continuity equations. It can be used for hydraulic modeling of flow through open canals and pipes using the Saint-Venant equations derived from the mass and momentum conservation (Haowen et al. 2020; Zhao et al. 2021). Equation (1) shows the continuity equation in each sub-basin.
formula
(1)
where dx is the water depth (m), t is time (s), P is the precipitation rate (m/s), E is the evapotranspiration rate (m/s); F is the infiltration rate (m/s), and q is water flow in each sub-basin per unit area (m/s) obtained by dividing the flow rate (Q) from the Manning equation (Equation (2)) by the surface area of the sub-basin (As).
formula
(2)
where W is the width of the sub-area (m), n is Manning's roughness coefficient, dp is the depth of pothole storage (m), and S is the slope of the sub-basin. Equation (3), which is the nonlinear reservoir equation, is obtained by substituting Equation (2) in Equation (1). The unknown parameter d can be obtained by solving Equation (3) using SWMM.
formula
(3)

Reports indicate that the sum of total concentration times of the individual basins and the concentration time required for water flow through the flooding canal to reach the outlet in Tehran basins has been less than 6 h (TehranMunicipality 2012; Movahedinia et al. 2019). Therefore, the model calculated rainfall-runoff correctly, and the local rainfall pattern was obtained based on the intensity-duration-frequency (IDF) curves using the alternating block method. Rainfall in Tehran is most severe (events longer than 3 h) in the middle of the precipitation period, i.e., from October to April. This was considered during the preparation of the report of the local rainfall pattern of the city. The hydraulics of the runoff was simulated for RPs of 10, 25, 50, and 100 years considering the critical role of floods in the design of hydraulic structures. The infiltration rates were determined using the Horton equations.

HEC-RAS simulation model

HEC-RAS has been widely used to perform one-dimensional, two-dimensional, or combined hydraulic calculations for natural and artificial canal networks and verify the flow state. The water surface profiles are achieved by solving the energy equation (Equation (4)) from one section to the next using an iterative algorithm (Barkhordari et al. 2020).
formula
(4)
where z1 and z2 are the levels of the main canal bed, y1 and y2 are the water depth in cross-sections, v1 and v2 are the mean velocity, α1 and α2 are the velocity coefficients, g is the acceleration of gravity, and he is the energy loss. Energy loss can be obtained by the summation of losses due to friction and the loss caused by opening and narrowing (Equation (5)).
formula
(5)
where is the slope of the energy line between two successive sections, C is the drop coefficient for opening and narrowing, and L is the weighted mean of the length of canal reach. The slope of the energy line at each point is obtained using the Manning formula (Equation (6)).
formula
(6)
where k is the section's transmission coefficient and can be calculated using Equation (7)
formula
(7)
where n is Manning's roughness coefficient, A is the section area, and R is the hydraulic radius.

Field studies and as-built plans of the canals cross-sections were used to determine cross-sections for the further development of the critical points in the HEC-RAS software. A broader area than the critical points was modeled because flow parameters, e.g., water surface profiles, are influenced by the upstream and downstream of the flow. Manning's roughness coefficients were between 0.015 and 0.021 based on the comprehensive plan of surface runoff. Nonetheless, the roughness coefficients were revised to 0.018 and 0.025 for the canal and floodplains based on experts' opinions. This was because erosion in the open canals affects the roughness coefficient, and the canals and floodplains are covered with cement and asphalt, respectively. Specific water level, the depth in the critical points, the normal depth, and the stage-discharge curves are used in the HEC-RAS model to define the boundary conditions. To introduce upstream and downstream boundary conditions, the normal depth based on the energy line slope of the canal upstream and downstream were used in this study. As an alternative, the bed slope of the canal can also be used with an adequate estimation if the energy line slope is not available. In this study, the bed slope of the three canals upstream and downstream was considered 0.01 based on the field survey. The geometrical properties of the bridge were also measured for the modeling. In order to control the hydraulic flow, characteristics such as flow velocity, flow depth, landing number, and water intake capacity were studied, regardless of the structures in the study interval, and longitudinal profiles for floods with different RPs in the canal were provided.

The performance of the stormwater collecting system in the current status

The SWMM software was used for the hydrologic and hydraulic simulation of the study area in its current status based on available information from field investigations. The observation data for the model calibration – measured by the operation staff and collected from Tehran municipality's civil technical deputy – includes the measured water level in nodes No. 5 and No. 15, and related to four storm events that happened in November 2019, five storm events in October 2020, and three storm events in April 2021, in which six events were employed for the calibration and the rest were used for the models' validation. The observed and simulated runoff depth were compared by the Nash-Sutcliffe Efficiency (NSE) and Root Mean Squares Error (RMSE). Within the different model parameters, the N-Impervious, N-Pervious, D-Store-Impervious, D-Store-Pervious, W, and CN were selected for the sensitivity analysis. The sensitivity analysis reveals that two parameters of CN and W variation showed the direct ratio with the total runoff (%), so the total runoff increased by increasing these parameters. Furthermore, within the two mentioned parameters, the sub-catchments width reflects the higher sensitivity values, so that with increasing 20% of this parameter, the total runoff increased by 1.72%. After the model's calibration, the validation process was performed based on the runoff depth for two individual storm events in November 2019, October 2020, and April 2021. The calculated indicator of NSE resulted in 0.803 and 0.792 in the calibration stage and 0.902 and 0.912 in the validation stage. Similarly, the RMSE indicator led to 1.55 and 1.03 cm in the calibration stage and 1.39 and 1.51 cm in the validation stage. Finally, the NSE and RMSE values in the calibration and validation stages verify the simulation accuracy.

The flood hydrograph obtained for a 25-year period was utilized to evaluate the capacity of the area. The storm sewer of the eastern flood diversion and the main canals leading to it collect and transfer mountain and urban catchment runoffs. Moreover, assessing and controlling the behavior of the flood control network under the current status in urban regions were performed considering 6 h rainfall with 10, 50, and 100-year RPs. The maximum allowable speed in the eastern flood diversion canal is 6 m/s, and it is covered with concrete. The hydraulic calculations for the current status and the flood with a 25-year RP in the study area showed that the main canal had insufficient flow capacity, and water overflew occurred in eight points. Comparing the inundation locations during floods with the simulation results confirmed these points' problematic nature. These nodes, which included bridges, were found from its northernmost to the southernmost points. Moreover, out of 207 km of the main network located in the eastern flood diversion catchment, about 54 km had insufficient hydraulic capacity. The results showed that the hydraulic power of the canal decreased and increased periodically over time.

The results of the hydraulic flood modeling with a 50-year RP at the problematic bridges were investigated to determine the most critical points. From north to south, these bridges were Khajeh Abdollah, Izadi, Janbazan, Hoseini, Golestan, Shora, Jafari, and Mahallati, with the discharge of 80, 136, 136, 162, 162, 167, 167, and 168 m3/s, respectively. The modeling results revealed that Izadi, Janbazan, Hoseini, and Jafari bridges have sufficient capacity in terms of size and discharge, while the rest showed a remarkable insufficiency in their capacity. The insufficiency in the Khajeh Abdollah Bridge was calculated at 75%, following the modeling results and showing land inundations. The insufficiency was estimated at 45 and 41% for the Shora and Mahallati bridges, respectively. Thus, the SWMM hydraulic modeling identified the critical bottlenecks along the eastern flood diversion floodway. The three bridges of Khajeh Abdollah, Shora, and Mahallati were considered for subsequent calculations due to their critical conditions, which are depicted in Figure 3.
Figure 3

The results of hydraulic flood simulation using SWMM, with RPs of 10 (a, b, c), 25 (d, e, f), 50 (g, h, i), and 100 years (j, k, l) for the critical points of the study area.

Figure 3

The results of hydraulic flood simulation using SWMM, with RPs of 10 (a, b, c), 25 (d, e, f), 50 (g, h, i), and 100 years (j, k, l) for the critical points of the study area.

Close modal

The dominant perspective of the City's experts and the decision makers for proposing the rehabilitation plans relies on their observations, mostly concluded from the inundation reports and statistics. Applying the stormwater simulation model of SWMM for simulating the hydraulic behavior of the main canal systems and the hydraulic conveyance structures in its path under precipitation with the different return periods shows the opposite outcomes in a couple of cases compared to the inundation reports. For instance, there were several node flooding reports for the Janbazan and Jafari bridges in the recorded historical data. However, the simulation results in the present study show different hydraulic behaviors for these two specific cases. These two contradictions forced the researchers of this study to examine the simulation results and the historical data more carefully and in more detail. The investigation of the lateral networks in the vicinity of both Janbazan and Jafari bridges and operational performance appraisal of the creeks and tributaries drained to the main canal at the upstream side of the bridges were the primary reasons for the inundation reports in both areas. Comparing the date of flooding problems reported by the citizens living within these two regions by the annual maintenance reports shows that the complaints are restricted to those years with a lack of financial resources for performing the maintenance activities, especially the dredging. Therefore, in contrast to the decision-making procedure in the status quo, which is mainly based on observations, the SWMM simulation model recognized the critical points based on the hydraulic behavior of the stormwater collecting system. Therefore, in light of the above, the SWMM simulation model, as mentioned in the study's objective, acts as a reliable decision support system for the managers and decision-makers and may reduce the subjectivity and uncertainty originating from the experts' perspectives.

Results of the HEC-RAS simulation model

The hydraulic flow simulation for the critical points was performed with 10, 25, 50, and 100-year RPs to investigate the hydraulic performance of the critical bottlenecks more accurately. The possible renovation and rehabilitation options for the critical bridges were discussed with urban flood management experts of Tehran municipality in brainstorming sessions based on field investigations (Figure 4). Field visits showed that there were four main causes of damage in these points: (a) lack of attention to hydraulic design, execution, and maintenance; (b) unsuitable locating according to the upstream and downstream winding course; (c) structural and geotechnical miscalculations for the bedding depth because of neglecting erosion and scouring; and (d) maintenance flaws. Following recommendations raised by experts in brainstorming sessions, the flow for each bridge was simulated using the HEC-RAS software to ensure that the bridges were critical based on the 10, 25, 50, and 100-year RPs. Therefore, field measurements were performed to estimate the course roughness, survey the canals upstream and downstream, and measure the bridge sizes.
Figure 4

The upstream and downstream sections of the three critical nodes: Khajeh Abdollah (first row), Shora (second row), and Mahallati (third row).

Figure 4

The upstream and downstream sections of the three critical nodes: Khajeh Abdollah (first row), Shora (second row), and Mahallati (third row).

Close modal
The hydraulic behavior of the critical canal reaches was simulated for a 50-year RP rainfall event using HEC-RAS. Figure 5 shows the hydraulic modeling results, i.e., water surface profiles before and after redesigning. Water surface profiles in the current status (Figures 5(a), 5(c) and 5(e)) reveal the canal's failure to pass the design floodwater in the three critical points, similar to the SWMM hydraulic modeling results.
Figure 5

The results of hydraulic flood simulation using HEC-RAS in current status (a, c, e) and after redesign (b, d, f) for the critical bottlenecks of Eastern Tehran.

Figure 5

The results of hydraulic flood simulation using HEC-RAS in current status (a, c, e) and after redesign (b, d, f) for the critical bottlenecks of Eastern Tehran.

Close modal

According to Figure 5(a), the local constriction along the Khajeh-Abdollah bridge, which included a reduction in width from 11 to 7 m upstream and to 4 m at the bridge, has reduced the cross-sectional area, in conjunction with the backflow and hydraulic potential reduction of the canal. This reduced the cross-sectional area and, therefore, flood passage capacity. As a result, flooding has occurred upstream in this bridge in the amounts of x, 2x, 3x, and 4x m for rainfall with RPs of 10, 25, 50, and 100 years, respectively. Figure 5(b) and 5(c) depict the same flooding pattern, indicating that the flooding occurred around x, 2x, 3x, and 4x m for rainfall with RPs of 10, 25, 50, and 100 years upstream. Various solutions were proposed in brainstorming sessions to modify the lack of flow capacity. Finally, three general solutions were considered: (a) deepening the canals, widening the canals and their combination; (b) increasing the bridge span; and (c) constructing a side drain. General indicators such as simplicity of implementation, traffic interruption, and flexibility in the face of varied installation issues or the foundation of nearby structures were investigated to evaluate and prioritize the proposed solutions for rehabilitating the critical points.

According to hydraulic calculations for the Khajeh Abdollah Bridge, the canal's cross-section should be increased by about 4.5 m to pass the flood with 50-year RP, at which it will collide with the 1.8-m sewer pipeline in the middle of the street. Therefore, the option of deepening and widening the flood control canal was chosen as the best alternative. By increasing the canal's depth by correcting the floor slope and installing a vertical drop upstream, the width of the bridge span increased from 4 to 7 m, equal to the upstream canal's width. The canal slope is assumed to be 0.0025 to control the velocity and flow regime (i.e., Froude number) in a flood with a 50-year RP.

For the Shora and Mahallati bridges, the option of widening and deepening was rejected for various reasons, including traffic disruption in the area and street closures adjacent to the canal, operation problems and diversion into the side drain, high costs, and execution time. To eliminate the hydraulic insufficiency in these two bridges, increasing the cross-sectional area of the bridge by increasing the canal's depth by 3 and 3.6 m, respectively, and adding a drop over the canals were selected as the most appropriate alternative. The canal slope near the Shora and Mahallati bridges should be 0.003 and 0.002, respectively, based on environmental conditions and the minimum dimensions required to repair the bridge bottleneck in the flood with a 50-year RP.

Figure 5(b), 5(d) and 5(f) depict the simulated water surface profile after redesigning all three critical nodes. Comparing water surface profiles before (a, c, e) and after local redesign (b, d, f) demonstrates that it is possible to pass floods with different RPs in these nodes by modifying the bridge's span. The average flow rate (Vave) in flood with an RP of 50 years is presented in Table 1 based on the environmental conditions and the minimum dimensions required to modify the bottlenecks and the canal slope. The table demonstrates that Vave in flood with an RP of 50 years is less than 5.25, 5.50, and 5.60 m/s in the Khajeh Abdollah, Shora, and Mahallati bridges, respectively.

Table 1

Hydraulic flow constraints in the vicinity of critical bottlenecks with a 50-year RP

Khajeh-AbdollahShoraMahallati
 Status quo 
Hydraulic constraints (max) VAve (m/s) 9.10 9.79 6.01 
Froude # 3.02 2.10 0.96 
 After rehabilitation 
Hydraulic constraints (max) VAve (m/s) 5.25 5.50 5.60 
Froude # 1.01 1.01 1.01 
Khajeh-AbdollahShoraMahallati
 Status quo 
Hydraulic constraints (max) VAve (m/s) 9.10 9.79 6.01 
Froude # 3.02 2.10 0.96 
 After rehabilitation 
Hydraulic constraints (max) VAve (m/s) 5.25 5.50 5.60 
Froude # 1.01 1.01 1.01 

Moreover, there is a risk of destruction and erosion during most of the operation. As shown in Table 1, the maximum Froude number for the 50-year flood on each bridge is 1.01. Therefore, according to the proposed standard values for these parameters by the USDA, necessary safety measures against erosion and destruction in the canal during the flood should be provided.

Results of the SWMM simulation model after rehabilitation

The dimensions of the critical points were substituted with new calculated values to investigate the performance of the proposed rehabilitation technique in this study. A new SWMM model was developed for the study area, and the simulations were conducted for rainfall with 10, 25, 50, and 100-year RPs. The floodwater flow at the critical bridges was investigated to study the effects of local rehabilitation. Figure 6 depicts the effects of canal rehabilitation on hydraulic performance at critical points, while Table 2 describes the hydraulic performance of the network before and after applying the optimal dimensions. The local rehabilitation of the eastern stormwater canal reduced the inundations in rainfall with various RPs. For floods with 10-year RPs, all the number of node flooding, inundated areas, and overflow volume decreased from 4, 17, and (10–45)% to 0. Similar results were obtained at higher RPs. After the rehabilitation, node flooding for floods with 25, 50, and 100-year RPs decreased from 10, 12, and 20 to 0, 2, and 4, respectively. Furthermore, implementing the local rehabilitation in the stormwater canal reduced the overflow volume from (30–65), (43–74), and (70–92) in the current status to (4–12), (11–27), and (24–36) for rainfall with 25, 50, and 100-year RPs.
Table 2

The hydraulic performance of the network before and after applying the optimal dimensions using SWMM

T = 10
T = 25
T = 50
T = 100
RPs (years)Node floodingFlooded area (%)Overflow volume (%)Node floodingFlooded area (%)Overflow volume (%)Node floodingFlooded area (%)Overflow volume (%)Node floodingFlooded area (%)Overflow volume (%)
 Current status 
Main canal 17 10–45 10 23 30–65 12 31 43–74 20 43 70–92 
 After rehabilitation 
Main canal 4–12 12 11–27 18 24–36 
T = 10
T = 25
T = 50
T = 100
RPs (years)Node floodingFlooded area (%)Overflow volume (%)Node floodingFlooded area (%)Overflow volume (%)Node floodingFlooded area (%)Overflow volume (%)Node floodingFlooded area (%)Overflow volume (%)
 Current status 
Main canal 17 10–45 10 23 30–65 12 31 43–74 20 43 70–92 
 After rehabilitation 
Main canal 4–12 12 11–27 18 24–36 
Figure 6

The results of hydraulic flood simulation using SWMM after optimizing, with RPs of 10 (a, b, c), 25 (d, e, f), 50 (g, h, i), and 100 years (j, k, l) for critical bottlenecks of Eastern Tehran.

Figure 6

The results of hydraulic flood simulation using SWMM after optimizing, with RPs of 10 (a, b, c), 25 (d, e, f), 50 (g, h, i), and 100 years (j, k, l) for critical bottlenecks of Eastern Tehran.

Close modal

Regarding the operational performance appraisal results of the stormwater collecting system before and after applying the recovery plan given in Table 2, it would be pointed out that the outstanding results concluded after rehabilitation with the assumption that regular inspections and maintenance activities are conducted. However, decreasing the maintenance measures' quality and quantity leads to higher flooding issues. For instance, the simulation results reveal that under the cases, the lateral networks work at 75% of their capacity during the storm events in rainfall with 25, 50, and 100-year RPs, the number of the nude flooding and inundation problems increased by 28, 35, and 41%, respectively. Furthermore, by increasing the clogging issues of the lateral stormwater collecting systems to 50%, the performance of the main drainage systems decreased to 30% in reducing the node flooding issues. Within the indicators calculated and reported in Table 2, the node flooding one is more important for the higher level of the city's managers, and the simulation results show that prioritizing the maintenance activities based on the critical points recognition strategy – proposed in this study – resulted to minimum flooding node within the network. In other words, under a specific scenario in which the tributaries that are located upstream of the critical bridges have zero clogging problems and the remaining lateral system work at 75% of their capacity, the number of nude flooding problems increased by 3, 8, and 17%, respectively for rainfall with 25, 50, and 100-year RPs.

This study introduces a practical approach to performing time-effective rehabilitation of stormwater drainage canals in urban areas. The proposed solution combined structural modifications that only included the localized improvement of critical points. The step-by-step approach used in this study to rehabilitate the structures distinguishes it from former rehabilitation projects for urban flood management. The critical points were determined in three steps: (a) hydraulic flow simulation of the Eastern Tehran floodway using SWMM to study the possible critical points in the design of six-hour rainfall with 10, 25, 50, and 100-year RPs; (b) field visits and brainstorming sessions with urban flood management experts to finalize the identification of critical points; and (c) field studies to measure physical-hydraulic parameters of the canal reaches in the vicinity of the critical points, and finally; (d) developing the hydraulic flow simulation model for each critical point in HEC-RAS to analyze the overflow more accurately.

Based on the numerous brainstorming sessions, various redesign options were analyzed for critical reaches, and deepening and widening were selected as rehabilitation solutions in the study area. As a result, the modified bed width was 7, 8.3, and 9 m for the critical periods, i.e., Khajeh Abdollah, Shora, and Mahallati bridges, and the modified depth was obtained at 3.4, 6, and 5.3 m, respectively. The new dimensions observed the study area's hydraulic and urban constraints. On average, the rehabilitation reduced inundation throughout the study stormwater canal by 46, 38, 39, and 27.5% for rainfall with 10, 25, 50, and 100-year RPs.

There was a significant limitation in the structural widening range because of the dense texture of neighborhoods and urban areas east of the Tehran metropolis. Also, there were limitations in increasing the canal depths because of the high gradient in the stormwater canal parts. Moreover, severe financial constraints of Tehran municipality limited the rehabilitation options. Various field investigations helped remarkably reduce measurements' uncertainty, making the results more applicable. Using LID methods as complementary techniques with local rehabilitation methods is suggested to increase the performance of the runoff drainage network. LID methods can be implemented in public areas such as parks, gardens, alleys, and streets within the stormwater drainage subnetwork. Therefore, inundation is expected to decrease remarkably in the Eastern Tehran stormwater canal, reducing the inflow from the subnetwork. Accordingly, it is suggested to determine critical subnetworks using the method introduced in this research and implement LIDs by prioritizing problematic and critical sub-catchments in each subnetwork.

Moreover, it is proposed to employ a systematic multi-criteria decision-making approach as a systematic weighting mechanism to prioritize the experts' perspectives and criteria (dealing with the social, economic, and technical priorities), acquired during the brainstorming sessions and field visits. It would be helpful to minimize the subjectivity of the experts' perspectives, especially for the other test cases where the City's experts cannot confidently recognize the dominant criteria. Moreover, another recommendation would be employing the supervised pattern recognition methods and especially the classification methods (such as SVM, Bayesian Networks, or other machine learning methods) in the conditions that the objectives, patterns, and principle issues of the sub-catchments are thoroughly transparent and specified for the city's authorities. As a suggestion, the critical points recognition procedure can be divided into two levels: (i) employing the proposed SWMM simulation model to determine the technical criteria and then (ii) training the classifiers based on the simulation model results and specified patterns recommended by the city's experts. It is worth noting that for those study areas where data gathering, interviews and questionnaires completion, and brainstorming sessions are restricted due to budget and time limitations, employing the unsupervised pattern recognition methods, including the crisp clustering algorithm (like the K-means algorithm) and the fuzzy clustering ones (such as FCM, FGK, FGG approaches) are recommended.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.

Alaghmand
S.
,
bin Abdullah
R.
,
Abustan
I.
&
Eslamian
S.
2012
Comparison between capabilities of HEC-RAS and MIKE11 hydraulic models in river flood risk modelling (a case study of Sungai Kayu Ara River basin, Malaysia)
.
International Journal of Hydrology Science and Technology
2
,
270
291
.
Al-Zahrani
M.
,
Al-Areeq
A.
&
Sharif
H. O.
2017
Estimating urban flooding potential near the outlet of an arid catchment in Saudi Arabia
.
Geomatics, Natural Hazards and Risk
8
,
672
688
.
Barkhordari
S.
,
Shahadany
S. H.
,
Taghvaeian
S.
,
Firoozfar
A.
&
Maestre
J.
2020
Reducing losses in earthen agricultural water conveyance and distribution systems by employing automatic control systems
.
Computers and Electronics in Agriculture
168
,
105122
.
Bertilsson
L.
,
Wiklund
K.
,
de Moura Tebaldi
I.
,
Rezende
O. M.
,
Veról
A. P.
&
Miguez
M. G.
2019
Urban flood resilience – A multi-criteria index to integrate flood resilience into urban planning
.
Journal of Hydrology
573
,
970
982
.
Carling
P.
,
Villanueva
I.
,
Herget
J.
,
Wright
N.
,
Borodavko
P.
&
Morvan
H.
2010
Unsteady 1D and 2D hydraulic models with ice dam break for Quaternary megaflood, Altai Mountains, southern Siberia
.
Global and Planetary Change
70
,
24
34
.
Coombes
P. J.
&
Downs
G.
2015
A decade of observations of local water cycle management at the Buderim Escape project
. In
36th Hydrology and Water Resources Symposium
.
Engineers Australia Hobart
, Tasmania, Australia.
Darajatun
A.
,
Sutrisno
M.
&
Budianto
B.
2020
The reservoir porous as a solution to control the urban flood in Pagarsih Road Bandung
. In
IOP Conference Series: Materials Science and Engineering
.
IOP Publishing
, Bristol, UK, p.
012033
.
Erdbrink
C.
,
Krzhizhanovskaya
V. V.
&
Sloot
P. M.
2014
Free-surface flow simulations for discharge-based operation of hydraulic structure gates
.
Journal of Hydroinformatics
16
,
189
206
.
Farzin
S.
&
Valikhan Anaraki
M.
2020
Optimal construction of an open channel by considering different conditions and uncertainty: application of evolutionary methods
.
Engineering Optimization
7,
1
19
.
Haowen
X.
,
Yawen
W.
,
Luping
W.
,
Weilin
L.
,
Wenqi
Z.
,
Hong
Z.
,
Yichen
Y.
&
Jun
L.
2020
Comparing simulations of green roof hydrological processes by SWMM and HYDRUS-1D
.
Water Supply
20
,
130
139
.
Hsu
N.-S.
,
Huang
C.-L.
&
Wei
C.-C.
2013
Intelligent real-time operation of a pumping station for an urban drainage system
.
Journal of Hydrology
489
,
85
97
.
Hua
P.
,
Yang
W.
,
Qi
X.
,
Jiang
S.
,
Xie
J.
,
Gu
X.
,
Li
H.
,
Zhang
J.
&
Krebs
P.
2020
Evaluating the effect of urban flooding reduction strategies in response to design rainfall and low impact development
.
Journal of Cleaner Production
242
,
118515
.
Kvočka
D.
,
Falconer
R. A.
&
Bray
M.
2016
Flood hazard assessment for extreme flood events
.
Natural Hazards
84
,
1569
1599
.
Lázaro
J. M.
,
Navarro
J. Á. S.
,
Gil
A. G.
&
Romero
V. E.
2016
Flood frequency analysis (FFA) in Spanish catchments
.
Journal of Hydrology
538
,
598
608
.
Lee
E. H.
&
Kim
J. H.
2018
Development of a flood-damage-based flood forecasting technique
.
Journal of Hydrology
563
,
181
194
.
Lord
S. A.
,
Ghasabsaraei
M. H.
,
Movahedinia
M.
,
Shahdany
S. M. H.
&
Roozbahani
A.
2021
Redesign of stormwater collection canal based on flood exceedance probability using the ant colony optimization: study area of eastern Tehran metropolis
.
Water Science and Technology
84
,
820
839
.
Mel
R. A.
,
Viero
D. P.
,
Carniello
L.
&
D'Alpaos
L.
2020b
Optimal floodgate operation for river flood management: the case study of Padova (Italy)
.
Journal of Hydrology: Regional Studies
30
,
100702
.
Movahedinia
M.
,
Samani
J. M. V.
,
Barakhasi
F.
,
Taghvaeian
S.
&
Stepanian
R.
2019
Simulating the effects of low impact development approaches on urban flooding: a case study from Tehran, Iran
.
Water Science and Technology
80
,
1591
1600
.
Nguyen-Tien
V.
,
Elliott
R. J.
&
Strobl
E. A.
2018
Hydropower generation, flood control and dam cascades: a national assessment for Vietnam
.
Journal of Hydrology
560
,
109
126
.
Pinar
E.
,
Paydas
K.
,
Seckin
G.
,
Akilli
H.
,
Sahin
B.
,
Cobaner
M.
,
Kocaman
S.
&
Akar
M. A.
2010
Artificial neural network approaches for prediction of backwater through arched bridge constrictions
.
Advances in Engineering Software
41
,
627
635
.
Ranaee
E.
,
Shoushtari
M. M.
&
Quchani
S. R.
2009
The combination of HEC-Geo-HMS, HEC-HMS and MIKE11 software utilize in a two branches river flood routing modeling
. In
2009 Second International Conference on Environmental and Computer Science
.
IEEE
, New York, NY, USA, pp.
317
321
.
Rodríguez
J. F.
,
García
M. H.
,
Bombardelli
F. A.
,
Guzmán
J. M.
,
Rhoads
B. L.
&
Herricks
E.
2000
Naturalization of urban streams using in-channel structures
.
Building Partnerships
2,
1
10
.
Silva
F.
,
Bonuma
N.
&
Uda
P.
2014
Flood mapping in urban area using HEC-RAS model supported by GIS
. In
International Conference on Flood Management
, São Paulo, Brazil, p.
9
.
Tao
J.
,
Li
Z.
,
Peng
X.
&
Ying
G.
2017
Quantitative analysis of impact of green stormwater infrastructures on combined sewer overflow control and urban flooding control
.
Frontiers of Environmental Science & Engineering
11
,
1
12
.
Tehran Municipality
2012
Summary Report of Studies. In: The Master Plan for Surface Water Management, T.a.D.O.o.T.M. (Ed.), Tehran, Iran, p. 269
.
Xie
J.
,
Chen
H.
,
Liao
Z.
,
Gu
X.
,
Zhu
D.
&
Zhang
J.
2017
An integrated assessment of urban flooding mitigation strategies for robust decision making
.
Environmental Modelling & Software
95
,
143
155
.
Yin
H.
,
Zheng
F.
,
Duan
H.-F.
,
Zhang
Q.
&
Bi
W.
2021
Enhancing the effectiveness of urban drainage system design with an improved ACO-based method
.
Journal of Hydro-Environment Research
38
,
96
105
.
Zhao
G.
,
Wan
Y.
,
Lei
Z.
,
Liang
R.
,
Li
K.
&
Pu
X.
2021
Effect of urban underlying surface change on stormwater runoff process based on the SWMM and Green-Ampt infiltration model
.
Water Supply
21, 4301–4315.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence (CC BY 4.0), which permits copying, adaptation and redistribution, provided the original work is properly cited (http://creativecommons.org/licenses/by/4.0/).